Dengue viruses (DENVs; serotypes 1-4) severely impact human health in all tropical and subtropical regions of the world, infecting 50-100 million people annually. DENVs are cycled between humans and their principal mosquito vector, Aedes aegypti. Populations of Ae. aegypti differ enormously in their vector competence (VC), a term that encompasses their ability to become infected by a virus, propagate the virus and ultimately transfer the virus to a vertebrate host. Variance in VC for DENV among Ae. aegypti populations is a complex quantitative genetic trait controlled by four principal factors: environment, mosquito genetics, DENV genetics and mosquito genetic by DENV genotype interactions. Quantitative genetic techniques showed that environmental factors account for ~60% of the variation in VC and thus only ~40% of VC variation is attributable to mosquito genetic effects. We used quantitative trait locus (QTL) mapping to show that many different mosquito genome regions control VC for DENV. We and many others have documented that different serotypes and genotypes of DENV affect VC in a given mosquito strain. This proposal focuses on the fourth principal factor: interactions between Ae. aegypti genes and alleles and DENV genomes. The leading mosquito candidates for this interaction are genes in the exogenous small interfering RNA (exo-siRNA) pathway. These genes detect the presence of viral dsRNA, dice up the viral genome, incorporate diced viral RNA as guide strands into the RISC complex which then slices and destroys the invading viral RNA genome. We have shown that exo-siRNA genes are among the fastest evolving genes in the Ae. aegypti genome, apparently due to an evolutionary arms race between mosquito innate immunity and viruses and/or transposable elements. We have shown that suppression of dicer2 expression via RNA interference (RNAi) greatly increases DENV replication in Ae. aegypti. The proposed work will test whether natural variation in key genes in the exo-siRNA pathways in Ae. aegypti interact with the DENV genome to select directly for novel DENV2 quasispecies. We will breed Ae. aegypti isofemale lines (AaILs) that exhibit statistically significant interactions in viral titers in the head with differnt DENV2 genotypes (D2gs) (Aim 1.1). For example, we will breed AaILs that exhibit high head titers when infected with one D2g (e.g. D2g-1) but exhibit low head titers when infected with a different D2g (e.g. D2g-2) and other AILs that have low head titers with D2g-1 but have high head titers with D2g-2. QTL mapping with Bulk Segregant Analysis (BSA) using the AaILs and D2gs from Aim 1.1 and resequencing technology based upon the recent availability of 9,074 physically mapped genes containing 163,468,873 nucleotides throughout the Ae. aegypti genome (Aim1.2) will test whether specific alleles in exo-siRNA genes cosegregate with the head titer with different D2gs. However, examination of mosquito genes only provides insights into half of this evolutionary arms race. We will also use AaILs and D2gs generated under Aim 1 to determine (Aim 2.1) whether viRNA profiles from mosquitoes with different VC vary, (Aim 2.2) define the impact of serial passage in defined AAILs on DENV-2 variation and (Aim 2.3) determine whether passage in one AAIL influences virus fitness in homologous and heterologous AAILs.